Kindle Sender Skill | Agent Skills

Kindle Sender

What This Does

Converts markdown documents from your Zettelkasten to EPUB and emails them to your Kindle for offline reading.

Three modes:

Research mode - Send all Research Swarm docs linked in today's daily note
Review mode - Send your Someday-Maybe.md backlog for Kindle review
Deep reading mode - Send with --expand-links to include full content of all linked documents as an appendix

When to Use

"Send my research to Kindle"
"Send today's research to Kindle"
"Send someday-maybe to Kindle"
"Send someday-maybe to Kindle with links expanded"
"Kindle my research"

Prerequisites (One-Time Setup)

Before first use, configure these environment variables in ~/.zshrc:

export KINDLE_EMAIL="your-name@kindle.com"
export GMAIL_USER="your-email@gmail.com"
export GMAIL_APP_PASSWORD="your-app-password"

Setup steps:

Find your Kindle email:
- Go to Amazon → Manage Content & Devices
- Click "Preferences" tab
- Look for "Personal Document Settings"
- Your Kindle email ends in @kindle.com
Whitelist your sender email:
- Same page, under "Approved Personal Document E-mail List"
- Add your Gmail address
Create Gmail app password:
- Go to Google Account → Security
- Enable 2-Step Verification if not already
- Go to "App passwords" (search for it)
- Create a new app password for "Mail"
- Use this password, NOT your regular Gmail password
Install Python dependency:
```
pip3 install pypandoc pypandoc_binary
```
Reload shell:
```
source ~/.zshrc
```

Instructions

For "Send my research to Kindle"

Find today's daily note:

/Users/eddale/Documents/COPYobsidian/MAGI/Zettelkasten/YYYY-MM-DD.md

Extract Research Swarm links from Captures section: Look for wikilinks matching [[Research Swarm - *]] pattern.

Run the sender script in dry-run mode first:

cd /Users/eddale/Documents/GitHub/powerhouse-lab/skills/kindle-sender/tools
python3 send_to_kindle.py --daily-note "YYYY-MM-DD" --dry-run

Show the user what will be sent: Display the list of documents found.
Ask for confirmation: Use AskUserQuestion: "Send these X documents to Kindle?"

If confirmed, run without dry-run:

python3 send_to_kindle.py --daily-note "YYYY-MM-DD"

Report results: "Sent X documents to your Kindle. They should arrive within a few minutes."

For "Send someday-maybe to Kindle"

Run the sender script with someday-maybe flag:

cd /Users/eddale/Documents/GitHub/powerhouse-lab/skills/kindle-sender/tools
python3 send_to_kindle.py --someday-maybe --dry-run

Show preview and ask for confirmation.

If confirmed, send:

python3 send_to_kindle.py --someday-maybe

For "Send someday-maybe to Kindle with links expanded"

Run with --expand-links flag for dry run:

cd /Users/eddale/Documents/GitHub/powerhouse-lab/skills/kindle-sender/tools
python3 send_to_kindle.py --someday-maybe --expand-links --dry-run

Show preview with list of documents that will be expanded.
Ask for confirmation, noting the expanded docs.

If confirmed, send:

python3 send_to_kindle.py --someday-maybe --expand-links

For specific files

User can also request specific files:

python3 send_to_kindle.py --files "Research Swarm - Topic.md" "Another Doc.md"

Script Options

| Flag | Description | |------|-------------| | --daily-note YYYY-MM-DD | Send Research Swarm docs from that day's note | | --someday-maybe | Send the Someday-Maybe.md file | | --files FILE1 FILE2... | Send specific markdown files | | --expand-links | Include content of linked docs (wikilinks) as appendix | | --dry-run | Show what would be sent without sending |

Troubleshooting

| Problem | Solution | |---------|----------| | "KINDLE_EMAIL not set" | Add to ~/.zshrc and run source ~/.zshrc | | "Authentication failed" | Use app password, not regular Gmail password | | "Document not delivered" | Check sender is whitelisted in Amazon settings | | "pypandoc not found" | Run pip3 install pypandoc pypandoc_binary |

Examples

Example 1: Morning research review

User: Send my research to Kindle
Claude: Found 3 Research Swarm documents in today's Captures:
  1. Research Swarm - Markdown to Kindle Workflow - 2026-01-09.md
  2. Research Swarm - API Design Patterns - 2026-01-09.md
  3. Research Swarm - Newsletter Growth - 2026-01-09.md

  Send these to your Kindle?
User: Yes
Claude: Sent 3 documents to your Kindle. They should arrive within a few minutes.

Example 2: Review backlog

User: Send someday-maybe to Kindle
Claude: Will send Someday-Maybe.md (your parked projects and tasks backlog).
  Send to Kindle?
User: Yes
Claude: Sent Someday-Maybe.md to your Kindle.

Example 3: Deep reading with expanded links

User: Send someday-maybe to Kindle with links expanded
Claude: Will send Someday-Maybe.md with 13 linked documents expanded:
  - PROJECT - Email Agent
  - Research - Claude Code Testing Infrastructure - 2026-01-06
  - ...

  The EPUB will include an appendix with the full content of each linked doc.
  Send to Kindle?
User: Yes
Claude: Sent Someday-Maybe.md (with 13 referenced documents) to your Kindle.

Agent Skills: Kindle Sender

Skill Files